768 ON THE ANCIENT ROUTE ALONG THE KONCHE-DARYA [Chap. XXI
trees along its banks lay on the ground in the form of shapeless splintered piéces of timber. Then the dunes became lower, living tamarisk bushes appeared amidst the dead tamarisk-cones, and after having covered a march of a little over sixteen miles we dropped rather suddenly upon a big winding lagoon of fresh water apparently left by the flooded Konche-daryd after the recent melting of its ice sheet. The river-bed, marked by rows of fine living poplars, was in sight about a mile off to the west. Here we camped when the camels came up at nightfall.
March to Next morning our parties divided, Afrâz-gul with the couple of Tikenlik men moving south
N. of along the Konche-daryd to the ferry of Konche-örtang (Turfân-karaul), while I turned to the
daryâ. north, in which direction lay the ruined ` Kurghan' according to Dr. Hedin's map. After pro-
ceeding about three-quarters of a mile, we crossed a dry river-bed of no great width running eastwards. Most of the dead trees lining it still stood upright. The ground beyond was covered with dunes from 6 to io feet in height ; among them dead tamarisk-cones were fairly frequent, proof that moisture from the river had reached here in the distant past. After a march of six miles we came, in fact, upon a large and perfectly marked river-bed, about 120 yards wide where we crossed it and about 8 feet deep, running from NW. to SE. From its direction it seemed very probable that it connected with the wide bed passed on the previous day's march. Its banks were lined with dead Toghraks, many fine trunks lying prostrate on the ground, while other smaller ones still stood upright. The gravel found at the bottom suggested that we were nearing the foot of the Sai.
Ancient bed Beyond this bed we crossed an expanse of bare clay overrun in places by light dunes ; their
linking with axis stretched from east to west, indicating that northerly winds prevailed in this region. Passing Kuruk-
daryà. several small beds which had no dead trees on their banks and looked as if they were formed only
by occasional drainage, we came upon living scrub and then, at a distance of some nine miles from camp, reached an imposingly wide bed lined with rows of big poplars, all fallen and much splintered. Their appearance suggested that they had died at a far more remote period than the trees found along the dry beds to the south. The bed was about 15o yards wide, while the rows of dead Toghraks along it formed a belt fully 50o yards across. The bed came from the north-west and stretched away in the direction of Ying-p`an. Judging from the relative positions shown by the plane-table, it seemed to me to lie in the direct continuation of the uppermost Kuruk-daryâ where this is met by the terminal flood channels of the Shindi river. To my regret lack of time and the limited water-supply available in our two tanks prevented me from following this obviously ancient bed right down to the point where alluvial deposits from those channels have probably either obliterated or deflected it. From a fairly high sand ridge flanking this bed we sighted the ruined ` Kurghdn ' through the haze to the NNW. and reached it after a march of about two miles from the point where we had come upon the ancient bed. The intervening tract yielded an abundance of hardy scrub, nourished, no doubt, by such occasional drainage as descends from the bare gravel Sai. Reeds, too, appeared in clumps close to the ruins, and suggested that if a well were dug here, water might perhaps still be reached at no great depth.